(no subject)
Feb. 19th, 2022 07:51 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
kradeelavkernel of a thought that probably won't make it in the identity zine, but is a topic I wish was talked about more in long-form:
I recently requested my data from amazon to see how much they've collected on me, and it's striking to see this consistent pattern of them - and other FAANG companies - always labeling me as a 18-34 "male" audience category (presumably because of stereotypical cis-male interests like tech/linux, videogames, planes, guns, etc). I do enough 'data camouflage-ing" that they usually don't get more accurate than that, but.
there's something tooth-gratingly offensive about that, to me, on several levels - like,
1) just the obvious "putting infinitely complicated souls" into boxes for capitalism, just to squeeze out a few more cents per person via advertising. particularly jarring when you remember all the advertising dollars put into pride month n' shit.
2) how dare these Big Data tech companies and algorithims assume that is my gender, moreover perpetuate this 1960's approach of gender = a narrow subset of interests? I don't identify as butch, but what if I was? what if I was a trans lady or nonbinary with those interests? for the wokescolds that a lot of these companies have turned into, this feels awfully regressive just on a basic "interfacing with the companies in a real world experience" level and not a PR or HR hypothetical.
3) I vaguely recall the mini-twitter-the-company scandal where they assigned similar advertising gender categories "as" you, and misgendered basically everyone that didn't fit the algorithms suspicions. it was morbidly funny when there was a blank and you could put "fuck you" as a gender (I did) as their hail mary loophole, but ... just, jesus.
4) it gets super black-mirror-y when you also realize just how much more algorithims are being used in medical data, credit score ratings, etc etc, not just a level of "let's get more money from them hurr hurr' but actually to the point of this-impacts-your-life-irrevocably-and-y-u-don't-even-know. bringing in IBM and the nazis is going to sound super alarmist/tin-foil-hat-y but I will never forget my holocaust class in college where the prof (i loved that dude) drilled in us that basically all of that shit wouldn't have happened without the (admittedly cruder) form of data collecting. where does it end?
anyway I'm getting huge helicopter-story-vibes of this big-data-slash-military-industrial-complex fucking with humanity that is literally by definition always going to be un-definable by algorithims.
I recently requested my data from amazon to see how much they've collected on me, and it's striking to see this consistent pattern of them - and other FAANG companies - always labeling me as a 18-34 "male" audience category (presumably because of stereotypical cis-male interests like tech/linux, videogames, planes, guns, etc). I do enough 'data camouflage-ing" that they usually don't get more accurate than that, but.
there's something tooth-gratingly offensive about that, to me, on several levels - like,
1) just the obvious "putting infinitely complicated souls" into boxes for capitalism, just to squeeze out a few more cents per person via advertising. particularly jarring when you remember all the advertising dollars put into pride month n' shit.
2) how dare these Big Data tech companies and algorithims assume that is my gender, moreover perpetuate this 1960's approach of gender = a narrow subset of interests? I don't identify as butch, but what if I was? what if I was a trans lady or nonbinary with those interests? for the wokescolds that a lot of these companies have turned into, this feels awfully regressive just on a basic "interfacing with the companies in a real world experience" level and not a PR or HR hypothetical.
3) I vaguely recall the mini-twitter-the-company scandal where they assigned similar advertising gender categories "as" you, and misgendered basically everyone that didn't fit the algorithms suspicions. it was morbidly funny when there was a blank and you could put "fuck you" as a gender (I did) as their hail mary loophole, but ... just, jesus.
4) it gets super black-mirror-y when you also realize just how much more algorithims are being used in medical data, credit score ratings, etc etc, not just a level of "let's get more money from them hurr hurr' but actually to the point of this-impacts-your-life-irrevocably-and-y-u-don't-even-know. bringing in IBM and the nazis is going to sound super alarmist/tin-foil-hat-y but I will never forget my holocaust class in college where the prof (i loved that dude) drilled in us that basically all of that shit wouldn't have happened without the (admittedly cruder) form of data collecting. where does it end?
anyway I'm getting huge helicopter-story-vibes of this big-data-slash-military-industrial-complex fucking with humanity that is literally by definition always going to be un-definable by algorithims.
(no subject)
Date: 2022-02-20 01:40 am (UTC)(no subject)
Date: 2022-02-20 09:34 pm (UTC)(no subject)
Date: 2022-02-20 01:55 pm (UTC)Kinda makes sense that our own deeply held (deeply masked?) biases would appear in our programs and robots.when you’re making something “predictive” you are going to be making a lot of assumptions, taking shortcuts (as the human brain tends to do, we love our labels and boxes much as we hate them and for now useless they are at holding the messy soup that is human experience), add marketing in the mix and it gets even worse.
Because successful marketing relies so heavily on prediction, patterns and maintaining a status quo; How can you sell the most of X when the consumer base is so varied. If you can’t put a group to it, it gets that much harder to tailor your convincing arguments as to why x belongs in your life. Diversity gums up the capitalist works, as they currently run. (You’d think diversity would just…open that consumer base a fuckton! That’s… so many more potential customers!)
How does one request this data? Also can any of this data be removed? I’d love to fuck up my recommendations on Amazon. |D
(no subject)
Date: 2022-02-20 09:33 pm (UTC)(ngl, sometimes i get ethically 'hmmmm' about *being* in marketing and seeing this shit up close but what keeps me here is it's a tangible product that is useful for folks, we don't data collect, and it's not some semi-scam-as-a-subscription-service.)
but damn, i also love your line about diversity gumming up capitalisim - not exactly "incompatible" with it, but it does certinally make it harder for folks to hack their way into near-infinite growth when it takes work to approach everyone as individuals. honestly feels like the good kind of checks and balances to me but that's a whole nother topic. xD loved hearing your thoughts here.
Here's a great article that goes in deleting it off of amazon specifically - https://lifehacker.com/how-to-download-everything-amazon-knows-about-you-its-1848412242 pretty sure it can't be removed
You may also want to talk to your folks about removing their info from whitepages dot com or mylife; I basically reverse-ddg my name and my parents' every year or so and send the sites takedown requests. they "have" to comply legally, but very often they'll get new data from a different person and eventually reupload. :v
(no subject)
Date: 2022-02-20 05:09 pm (UTC)I am curious about the connection between data collection and the holocaust that your prof emphasized-- it certainly makes intuitive sense, because how do you commit genocide without being able to categorize people into groups in the first place. Do you have any books or articles to recommend on the subject?
(no subject)
Date: 2022-02-20 09:41 pm (UTC)and absolutely to any articles/books there! idk about the quality of the overall site, but this felt like a pretty great "explain like i'm ten" article on the topic that matched up with what my prof was saying: https://allthatsinteresting.com/ibm-nazis-ww2
for a deeper dive, "IBM and the Holocaust: The Strategic Alliance between Nazi Germany and America's Most Powerful Corporation" is a book written by an investigative journalist and was a bestseller, so, reasonably decent quality there.
mind, it wasn't "just" IBM that was doing this as the prof also emphasized that the nazis basically imported american eugenics among other tech - but it's definitely the clearest link to pull on as a thread to explore more.
(no subject)
Date: 2022-02-21 04:05 am (UTC)I... don't think it was ever really an intentional thing to not use amazon? I mean, I guess it is now, in that I wouldn't choose to start at this point, but it just never really occurred to me. I don't like shopping online and never order anything that I can buy in a store, mostly because anything I intend to buy online I'll probably just forget about and procrastinate on until I no longer need or want it. There's no instant gratification to actually walking out of a store with the item, so I can't convince myself that buying stuff online will actually result in my ownership of the object 😂 And I know some people use it for... groceries and stuff? But I think its reach in Canada has always lagged enough behind its offerings in the US that it just never seemed convenient.
I think the only thing I've ever bought from it was a set of MCAT study books, which were expensive enough that the university bookstore told me to buy them on amazon instead of from them, and there didn't seem to be any other source, and my partner has an amazon account so I just asked them to buy it.
(no subject)
Date: 2023-01-14 08:38 pm (UTC)(no subject)
Date: 2023-01-24 11:31 pm (UTC)glad to see you on here too! <3